120 ALASKAN GLACIER STUDIES
changed, and the flat itself was unaltered excepting by being narrowed by advance of the ice on the inner side; but, all above the interior flat was completely changed. Far up the valley, to the point where we had so easily walked ten months before, the ice surface was a maze of jagged pinnacles and seracs with intervening crevasses. The ice rose higher in its valley. It extended to the walls of the mountain valley. All smooth ice was gone, and the medial and lateral moraines were destroyed. It had become an entirely different glacier in almost every respect in the brief interval of ten months. From the valley mouth to its head the glacier was impassable, and there was not a half acre that was unbroken. The entire valley glacier was broken as an ice fall is in Alpine valleys, or as the Hubbard Glacier is back of its ice cliff. It seemed almost incredible that ten months before one could go over this surface in any direction, encountering a crevasse only here and there. In August, 1905, sleds could have been easily drawn over the valley glacier; in 1906 one could go out on the glacier only with the greatest difficulty.
The crevassing extended throughout practically the entire inner bulb, destroying the series of colored moraines of 1905, and replacing this area of moraine-covered ridges and hummocks with a maze of jagged pinnacles and deep crevasses. In 1905 from a distance no ice could be seen in this area of banded moraines, but in 1906 fully half the surface was clear ice. Much of the moraine had fallen into the crevasses, and the only indication of the former series of banded moraines was the presence, here and there, of a faint tinge of color on the broken ice surface. Our attempts to climb upon the broken ice to a point where we could photograph from approximately the same position which we occupied in 1905 were baffled by the intricate maze of crevasses which extended, in all directions, to the very edge of the inner bulb.
The periphery of the inner bulb had been pushed forward a 100 yards or more, overriding the granite gorge through which the glacial torrent emerged upon the interior flat in 1905 (PL LVH, A,) and narrowing the interior flat somewhat. There was a notable thickening of the inner bulb also, but we had no means of determining the amount, though the front cannot have been less than 200 or 300 feet higher in 1906 than it was in 1905. The advance of the front of the bulb upon the interior flat was in the nature of overriding rather than a pushing forward of the ice beneath it, plainly proving that the ice beneath the flat was very thin. For this reason, the effect of the thrust was not communicated across the flat to the outer bulb, which remained the same in 1906 as it was in the previous year.
One of the most notable changes brought about by the advance was upon the marginal drainage. By overriding the granite gorge the outflow of the glacial torrent at this point was checked and, for the time being, the interior flat was robbed of this, the greatest source for the alluvial deposits that had been accumulating in it. At the same time, naturally, the volume of the glacial stream that outflowed through the ice gorge across the outer bulb was correspondingly diminished and, therefore, the principal source of water for the building of the western part of the great alluvial fan beyond the outer bulb was withdrawn. This alluvial fan, the largest in Eussell Fiord, was notable not only for its size but also because its eastern half was covered by an alder thicket, while its western half, crossed by the branches of the stream just mentioned, was barren of vegetation, as is normal for growing alluvial fans in this region. The presence of alder thickets on the eastern half of the fan was interpreted as proof of changes of earlier date. First there must have been streams flowing over it in order to build it; then, by some